Re: HDFS upgrade problem of fsImage

I insist hot upgrade on the test cluster because I want hot upgrade on the prod cluster. On 2013-11-21 7:23 PM, "Joshi, Rekha" wrote: > Hi Azurry, > > This error occurs when FSImage finds previous fs state, and as log states you > would need to either finalizeUpgrade or rollback to proceed.Bel

Build failed in Jenkins: Hadoop-Hdfs-0.23-Build #797

See -- [...truncated 7699 lines...] [ERROR] symbol : class Parser [ERROR] location: package com.google.protobuf [ERROR]

Hadoop-Hdfs-0.23-Build - Build # 797 - Still Failing

See https://builds.apache.org/job/Hadoop-Hdfs-0.23-Build/797/ ### ## LAST 60 LINES OF THE CONSOLE ### [...truncated 7892 lines...] [ERROR] location: class com.google.pro

Re: HDFS upgrade problem of fsImage

Thanks Steve, I'd like to contribute for this JIRA. so do you mean HDFSv2 doesn't support rolling upgrade currently? On Thu, Nov 21, 2013 at 6:16 PM, Steve Loughran wrote: > Given you are asking questions on hdfs-dev rather than hadoop-user, the > normal home of user-level questions, i'd like

Re: HDFS upgrade problem of fsImage

Given you are asking questions on hdfs-dev rather than hadoop-user, the normal home of user-level questions, i'd like to extend to you the opportunity to help the HDFS dev team to implement rolling updates https://issues.apache.org/jira/browse/HDFS-5535 design reviews, code and tests welcome -st

Re: Metrics2 code

Hi Andrew, thanks for your reply. The metrics2 configuration is : *.sink.ganglia.class=org.apache.hadoop.metrics2.sink.ganglia.GangliaSink31 *.sink.ganglia.period=10 *.sink.ganglia.supportsparse=true namenode.sink.ganglia.servers=10.232.130.6:8649 datanodenode.sink.ganglia.servers=10.232.130.6:864